Skip to content

YeongHyeon/MemAE-TF2

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

[TensorFlow 2] Memorizing Normality to Detect Anomaly: Memory-augmented Deep Autoencoder for Unsupervised Anomaly Detection

TensorFlow implementation of Memorizing Normality to Detect Anomaly: Memory-augmented Deep Autoencoder for Unsupervised Anomaly Detection. [PyTorch Version] [TensorFlow 1 Version]

Architecture

Architecture of MemAE.

Graph in TensorBoard

Graph of MemAE.

Problem Definition

'Class-1' is defined as normal and the others are defined as abnormal.

Results

Restoration result by MemAE.

Box plot and histogram of restoration loss in test procedure.

Environment

  • Python 3.7.4
  • Tensorflow 2.1.0
  • Numpy 1.18.1
  • Matplotlib 3.1.3
  • Scikit Learn (sklearn) 0.22.1

Reference

[1] Dong Gong et al. (2019). Memorizing Normality to Detect Anomaly: Memory-augmented Deep Autoencoder for Unsupervised Anomaly Detection. arXiv preprint arXiv:1904.02639.